parallelSimes: Combine parallel p-values with Simes' method

In LTLA/metapod: Meta-Analyses on P-Values of Differential Analyses

Combine parallel p-values with Simes' method

Description

Combine p-values from parallel tests with Simes' method. Each group of p-values is defined from the corresponding entries across all vectors.

Usage

parallelSimes(p.values, weights = NULL, log.p = FALSE)

Arguments

p.values	A list of numeric vectors of the same length, containing the p-values to be combined.
weights	A numeric vector of positive weights, with one value per vector in .... Each weight is applied to all entries of itscorresponding vector, i.e., all p-values in that vector receive the same weight. Alternatively, a list of numeric vectors of weights with the same structure as p.values. Each p-value is then assigned the weight in the corresponding entry of weights. Alternatively NULL, in which case all p-values are assigned equal weight.
log.p	Logical scalar indicating whether the p-values in p.values are log-transformed.

Details

The joint null hypothesis for each group is that all of the individual null hypotheses are true. Simes' method will reject the joint null if any of the individual nulls are rejected, providing weak control of the family-wise error rate.

In theory, the method is only applicable to independent tests, but experience suggests that it is quite robust to dependencies. The calculation itself is very closely related to the Benjamini-Hochberg method for controlling the false discovery rate. One can actually obtain Simes' combined p-value by taking the smallest BH-adjusted p-value across a group.

If non-equal weights are provided, they are treated as relative frequency weights. That is, if one p-value is given a weight of 10 and another p-value is given a weight of 1, the former is considered to occur 10 times more frequently than the latter.

The representative test for each group is defined as the test with the p-value that is ultimately used as the combined p-value. Briefly, one can identify this test as that with the smallest BH-adjusted p-value if the monotonicity adjustment were omitted. The influential tests for each group are defined as those with p-values no greater than the representative test's p-value. This is based on the fact that increasing them (e.g., by setting them to unity) would result in a larger combined p-value.

Value

A list containing:

• p.value, a numeric vector of length equal to the length of each vector in p.values. This contains the combined Simes p-value for each group, log-transformed if log.p=TRUE.

• representative, an integer scalar specifying the representative test in each group. Specifically, this refers to the index of the vector of p.values containing the representative test.

• influential, a list of logical vectors mirroring the structure of p.values. Entries are TRUE for any p-value that is deemed “influential” to the final combined p-value.

Author(s)

Aaron Lun

References

Simes RJ (1986). An improved Bonferroni procedure for multiple tests of significance. Biometrika 73:751-754.

Sarkar SK and Chung CK (1997). The Simes method for multiple hypothesis testing with positively dependent test statistics. J. Am. Stat. Assoc. 92, 1601-1608.

Benjamini Y and Hochberg Y (1997). Multiple hypotheses testing with weights. Scand. J. Stat. 24, 407-418.

See Also

groupedSimes, for a version that combines p-values based on a grouping factor.

Examples

p1 <- rbeta(100, 0.8, 1)
p2 <- runif(100)
p3 <- rbeta(100, 0.5, 1)

# Standard application:
out <- parallelSimes(list(p1, p2, p3))
str(out)

# With vector-level weights:
out <- parallelSimes(list(p1, p2, p3), weights=c(10, 20, 30))
str(out)

# With log p-values. 
out <- parallelSimes(list(log(p1), log(p2), log(p3)), log.p=TRUE)
str(out)

LTLA/metapod documentation built on Jan. 19, 2024, 11:49 p.m.